Celebrating 20 Years
Top 10 Find a Projector Reviews Throw CalculatorCalculator Buyer's Guide Projector Forums
Ease of Use
Intended Use:
Home Theater
Optoma HD91 Projector Optoma HD91
(add to Compare List)
Go to My Compare List

Street Price: n/a
3D: Full HD 3D
Weight: 15.4 lbs
Aspect Ratio:16:9
Lens:1.9x manual
Lens Shift:H + V
Lamp Life:20,000 Hrs
20,000 (eco)
Warranty:3 year
Connectors:  Composite, Component, VGA In, HDMI 1.4a (x2), USB, RS232, 12-Volt Trigger (x2),
Video Formats:  480i, 480p, 720p, 1080i, 1080p/60, 1080p/24

Review: Optoma HD91
LED 1080p Home Theater Projector

Bill Livolsi, March 26, 2014


Light output. The HD91 is rated at 1,000 ANSI lumens. After a 15-minute warm up, the projector's internal white test pattern gave a reading of 880 lumens, which was the maximum sustained output we obtained in our testing. Unlike lamp-based projectors, LED-based projectors reach their maximum brightness immediately upon startup. And whereas lamp-based models "warm up" to their full brightness over a few minutes, LED-based projectors like the HD91 actually lose brightness over the first few minutes until they stabilize.

The brightest usable image mode was the aptly named Bright mode, which measured 840 lumens. Bright mode emphasizes green, reduces contrast, and is locked to the DynamicBlack 3 power setting. It is suitable for max-brightness applications like watching sports in a room with some ambient light, though its low output means that ambient light should still be curtailed as much as possible to avoid washing out the image.

Cinema mode is the preferred operating mode for home theater use. At 517 lumens, Cinema mode offers much better color and contrast than Bright mode while still maintaining most of its light output. Photo mode, at 513 lumens, is similar to Cinema mode, though it contains small changes to gamma and color saturation that enhance the visibility of mid-tone detail. The remaining modes are Reference (408 lumens) and Film (386 lumens), both of which attempt to provide a more accurate image than Cinema mode but sacrifice light output to obtain it. Overall, we felt that the trade-off was not worth the loss in brightness, so we stuck with Cinema mode for most viewing.

The biggest brightness concern on the HD91 is the 1.9:1 zoom lens. At the telephoto end of the zoom, the lens reduces light output by 46%, bringing Cinema mode from 517 lumens to 278 lumens. Be aware of this reduction before mounting your projector.

Contrast. Out of the box, the projector's black level is a touch higher than some other home theater projectors, but some fine-tuning of the projector's gamma controls can improve black level significantly. Black level after calibration is comparable to many other home theater projectors. Contrast in any given scene, on the other hand, is particularly good. The picture showed above average pop, with great background separation and three-dimensionality.

Care should be taken when setting the projector's gamma controls, as the projector has excellent shadow detail once calibrated but doesn't look particularly impressive out of the box. Our preferred setting uses the 2.4 preset with a -3 offset. On our test unit, this gives depth to shadows without crushing detail and vastly improves the projector's contrast and shadow detail.

While DynamicBlack is primarily a brightness control, it also improves black level in dark scenes. While the choice of whether or not to use DynamicBlack is ultimately yours, we do offer this word of advice. If you plan to use DB, calibrate your projector with DB enabled. The DynamicBlack control alters the power levels of the LEDs differently, so a perfect calibration with DB turned off will undergo significant color shift if DB is then enabled later.

Color. The HD91's color controls are intuitive and comprehensive, so it is easy to calibrate the projector if you have the knowledge and equipment to do so. If you don't, you can hire a local AV professional to fine-tune your projector.

Sharpness and clarity. The HD91 has a crystal-clear HD image with plenty of detail, and that's without using the projector's UltraDetail sharpening system. With that system engaged, fine detail jumps out from both standard-def and HD content. The effect can actually be a little overwhelming if you're using a source that is already high in detail, such as The Dark Knight or Baraka on Blu-ray.

Input lag. After extensive testing, the HD91's lag performance the same in every image mode provided PureMotion is disabled. With PureMotion off, the HD91 measures 76 milliseconds of input lag, or about four and a half frames on a 60 fps signal. PureMotion increases lag to 126 milliseconds (7.5 frames).

Previous Page
Key Features
Next Page
Review Contents: The Viewing Experience Setup and Configuration Key Features Performance
  Limitations Shootout vs Epson 5030UB Conclusion

You support ProjectorCentral when you buy
the Optoma HD91 from these
authorized resellers.
Thank you!

Reader Comments(11 comments)

Posted Mar 27, 2014 3:07 AM PST

By crabarts231

Post a Comment Alert Moderator
If I want to turn the Dynamic Black on what Dynamic Black you guys suggest? DB1, 2, or3?? Of course I will calibrate with DB on.

Posted Mar 27, 2014 10:19 AM PST

By Robert

Post a Comment Alert Moderator
Bill, thanks for the detailed review.

One major benefit of LED projection is that brightness remains relatively unchanged over the long haul. So, once you set up and calibrate the projector, that's it.

With lamp based projectors, how much dimmer is the bulb after 100 hours, 500 hours or 1000 hours? Brightness of a new lamp tends to drop significantly and then stabilize over time. So, how long will it take for a lamp based projector to be less bright than the LED projector? Also, as the lamp fades, colors and black detail will change, requiring recalibration.

It would be interesting to compare measurements after long term testing - after 100, 500 and 1000 hours.

Posted Mar 28, 2014 7:57 AM PST

By Edgar_in_Indy

Post a Comment Alert Moderator
It would be a great selling point for LED-based projectors if they maintained their maximum brightness over the full life of the light source, but I have heard some reports from other LED projectors that they actually do lose brightness over time. So it will be interesting to see the long-term reports come in.

Posted Mar 31, 2014 2:44 PM PST

By Bill Livolsi

Post a Comment Alert Moderator
Hi everyone. Thanks for reading.

crabarts231 - it really depends on the content you're watching and the room you're watching it in. The higher levels of DynamicBlack will dim brightness more than the lower levels, so unless your room is dark and non-reflective enough to eliminate ambient light, the downsides may outweigh the benefits.

Robert - obviously this depends on the projector in question. I have yet to see a trustworthy source for long-term lamp dimming that is generalizable across all projectors, but it is clear that many projector lamps lose a chunk of brightness in their first 100 hours and then gradually fade from that point. With our specific comparison (Epson 5030UB versus Optoma HD91), I don't think the cross-over point exists.

The problem with long-term testing (and the reason why we haven't done something like this already) is that you need to maintain a normal pattern of use for your data to be meaningful. In other words, we can't just burn the lamp for 24 hours a day, 7 days a week, and expect to get results that are applicable to typical home use. Instead, we'd have to turn the projector on for smaller periods of time (say, 2 to 4 hours at a stretch), then turn it off and allow it to cool down before starting it up again. This takes a lot more time than just running the projector all day. If we used a "six hours on, six hours off" schedule on a projector with a lamp life of 4,000 hours, it would take just under a year to exhaust the lamp's life span.

Still, it is a topic that we are interested in, and we will likely write more about the differences between mercury lamps and alternative light sources in the future.

Edgar_in_Indy - To our knowledge, the projectors that suffer the largest light loss are hybrids -- those that use a combination of LED and laser sources. Pure LED projectors have less brightness but also lose less light output over their lives. Some estimates (unconfirmed) put total brightness loss of LEDs at 30% over the entire specified lifespan, and that loss is heavily weighted towards the end.

Posted Apr 4, 2014 12:25 PM PST

By Frank

Post a Comment Alert Moderator
What a great ittle projector! Of course, for LESS MONEY than what Optoma is peddling their HD91 for, you could get a PROFESSIONAL LED+laser light engine projector from Panasonic (Model PT-RZ470) that outputs 3500 ANSI lumen instead of just 1000. Just sayin'....

Posted Apr 4, 2014 2:03 PM PST

By Stunko

Post a Comment Alert Moderator
Funny that for a 1000 lumen PJ Optoma actually gives in their spec sheet a maximum projection size of 300-inches. Do the law of optics and physics support this size screen for any 1000 ANSI lumen projector, I wonder?

Posted Apr 4, 2014 2:10 PM PST

By Frank

Post a Comment Alert Moderator
"... projectors that suffer the largest light loss are hybrids -- those that use a combination of LED and laser sources. Pure LED projectors have less brightness but also lose less light output over their lives."

Trying to figure out how this goes. Okay, so a pure LED projector used R+G+B LEDs, whereas a hybrid projector uses a RED LED, a BLUE LED, and for the GREEN color, it uses a blue laser with a yellow phosphor wheel. If these LED+laser combination "hybrid" projectors lose more of their brightness than the pure LED chip projectors do, as seems to be suggested here, than the lightness loss must be related to the laser part in the hybrid, specifically the tiny blue laser gun that is used to generate the greens via the spinning phosphor wheel. I was not aware that laser sources dim over time at all, although I know that LEDs do.

Posted Apr 4, 2014 2:43 PM PST

By Bill Livolsi

Post a Comment Alert Moderator
Stunko - a projector emitting exactly 1000 lumens, projecting onto a 300" diagonal screen at 1.0 gain, would measure 3.8 foot-Lamberts. However, using a screen of equal size with a gain of 4.5 would give you 17 fL. So yes, it is possible.

Frank - I may have overstated things a little.

Some hybrid models we've tested experienced significant loss of light output over the first few hundred hours of runtime, specifically in the green channel. I'm not an engineer, but I suspect it was related to the phosphor. Then again, we tested this phenomenon years ago when the first Casio hybrid projectors were released. We did not see that same dramatic degradation in the Panasonic hybrid model we reviewed.

So far, we have not seen evidence of similar rapid degradation in projectors that are solely LED-based, but we have also only reviewed a small handful of pure LED projectors. This is something we'll be testing further as the market matures.

It's possible that the loss of green we observed in the early Casio hybrids was a model-specific problem that will not reappear in other hybrid models, but I was commenting based on my own experiences. I realize now that I phrased my comment too broadly, and for that I apologize.

Posted Apr 21, 2014 11:49 AM PST

By Art

Post a Comment Alert Moderator
Hi Bill, You done with that HD91 yet, I'm still trying to get one out of Optoma to review…

Great review, btw, I tend to agree, and do believe from my conversations with them and others, that their green problem has to a large degree tied to the design of their first (and second?) gen systems.

To Stunko, let's just say some manufacturers tend to be more optimistic than others. Interestingly 25 foot diagonal has been the favorite "maximum size" of most projectors going back to the early days. Back in 1994, the first "portable" the Proxima DP2800 projector claimed 110 lumens and claimed maximum screen size was 25' diagonal - a popular size used in a great many hotel ballrooms. Of course back then, all lights but emergency exits were normally out so that the screen was bright enough to see… Some times marketing trumps reality. -art

Posted May 27, 2014 12:33 PM PST

By Doug Blackburn

Post a Comment Alert Moderator
Calibrating with Dynamic Black ON is not going to be as easy as you might think -- I haven't thought about this in GREAT depth yet, but it seems to me that to get best results if you want to use Dynamic Black is to do calibration in 2 phases. This will be required to get decent results...

Phase 1 - Turn Dynamic Black off and calibrate everything as best you can

Phase 2 - Turn Dynamic Black ON to the setting you plan to use and as you fine tune grayscale steps and even colors, you will need to make your adjustments so that you do not change the luminance of the step you are adjusting. For example. Let's make this easy and say with DB OFF you calibrated the grayscale and you measured 10% White at 0.1 fL and you set your gamma to 2.25 (with variations in gamma that are practically inevitable, you managed to maintain your gamma between 2.2 and 2.3 for 10%-100% white). Now remember, this is with DB OFF. OK, so now you have a good grayscale and good gamma. Now, you turn DB ON and pick your setting/level (1-3). For whatever reason, when you measure some steps, especially darker steps, there are color shifts you need to get rid of. So now you measure 10% white as 0.05 fL due to the action of DB. In order NOT to screw up that gamma you set when DB was OFF (remember, your result where each grayscale step was between 2.2 and 2.3 gamma), you now need to remove some green from 10% white but when you do that, 10% white is going to get quite a lot dimmer... you don't want to let that happen. You need to make your 10% adjustments but you have to keep your luminance at 0.05 fL. So you can't just take out some green. If you were to determine that -4 on green would "fix" 10% white... your luminance level might drop to 0.04 or even 0.03... you can't let that happen. To keep that from happening, you might have to make your adjustments something like -1 for green and +3 for red and +3 for blue. That way you are removing green but also adding red and blue to keep that step from getting dimmer (these numbers are samples only, each projector is going to respond differently).

If you maintain luminance of each step after you have calibrated with DB off, you should maintain your gamma performance pretty darn well.

Similarly, when adjusting colors (red, green, blue, cyan, magenta, yellow), one of the things you SHOULD be looking at is color luminance. After you turn DB ON, your first measurement of each of the 6 colors will give you the target luminance for each color that you want to maintain as you make each of the 6 colors accurate. If your Red luminance was 3.36 fL, for example (with DB ON) but you need to tweak red to make it more accurate, you will want to do what you can to maintain that 3.36 fL of luminance which may mean making adjustments of multiple controls to keep your luminance at 3.36.

Doing this should allow you to use DB (or Auto Iris) while not introducing odd gamma behavior as image brightness goes up or down in response to the iris or changing light levels of the LED light source at different calibration settings. You may be used to having your d65 target (x=313, y=329), but luminance determines gamma -- so when adjusting any step AFTER the DB-OFF phase of calibration, you must maintain the DB ON luminance you began with... skipping that step is how we get those strange gamma results when using DB or Auto Iris.

Posted Jul 30, 2014 2:57 AM PST

By Manuel

Post a Comment Alert Moderator
Great report, though, for me one major positive of aspect of LED is the instant on and off feature, making a projector equal in speed to a TV and less hassle to start. We often use ours only 20 min/day, so this makes a difference. One question remains: why aren't there 2000 lumen LED home theater projectors yet?

Post a comment

Commenting on this article is easy and does not require any registration. Your email address is necessary for you to activate your comment once it has been submitted. It will not be shown to other site viewers. ProjectorCentral reserves the right to remove any comment at any time for any reason. Foul language is not permitted, nor are personal attacks. No HTML allowed. All comments should remain on topic.


Email Address:(used only to confirm your comment)

Your Comment:

(Enter the numbers as they appear to the left)